Cloud advocates want you to believe that you can resolve all your business challenges (let alone just your IT challenges) by moving to the cloud. Sure, as a buyer, you’re going to exercise a degree of skepticism; but if everyone’s doing it, how can you go wrong? Just like a real cloud journey, there are real possibilities for turbulence, sometime quite violent. During that period it may become quite terrifying. Fortunately, there is an extremely high probability that your organization (whether it is a vast enterprise or just a one-person show) will in fact survive the experience, but there will often be some bumps and bruises and it may cost you quite a bit to repair the damage or recover the customers lost during the transition.
The promise is that once you get up to altitude and past all the weather effects, your organizational processes and assets (in other words, the intellectual property that makes your organization what it is) will move faster, with better tools to manage them, and they will be handled properly. You won’t have to keep people trained up on the latest technology, and you’ll have built-in tools that help you make your processes move faster and/or more effectively. Best of all, if you find that your future solutions require hosting one layer or another of your solution elsewhere, then whoosh! You can simply switch it to another cloud provider, and you can even do this one layer at a time. Software as a Service, Platform as a Service or Infrastructure as a Service, or all the above, or some of the above. A few clicks of a mouse and you’re up and running in the new cloud in no time at all having captured whatever savings may be offered.
Peridot Solutions, my 8-6 employer, is conducting several engagements with commercial companies to help them work this magic, and the reality is that there is no magic in it. It’s a complex undertaking. It doesn’t have to take forever and it doesn’t have to cost millions of dollars to make the shift, but it does require a lot of thinking about how to do it and quite a bit of work to make it work right … and that’s if you’re already using or at elast prepared for a cloud service, i.e. your solutions are configured in some sort of organized manner instead of being cobbled together in ways that only your particular band of gnomes are aware of.
At a much simpler level, if you’re a subscriber to my highly-occasional email updates you know that I recently announced a switch in email responders. The size of my contacts list makes this a tiny operation by almost any standard, so how hard could this be? Well, let’s just say it brought a couple of the known advantages of cloud operations into sharp relief:
- Have it your way: Well, no. The cloud provider delivers reliable service and makes money by doing a lot of things the same way. You have to do it their way. you just get to pick from a few different flavors of their way which makes it seem as though you are exercising control, but what you are really exercising is a choice among a limited number of options.
- The deal is the deal: Well, no. One is that your service provider can change the rules on you. This mostly applies to smaller operations that are service-takers and price-takers. My original provider was riding on a Google-provided platform. Unknown and unannounced to me, and most certainly without consultation, Google decided on new limits to the amount of mail you can send from your own email address. Not only is that annoying to me, but it will drastically reduce the intervening provider’s ability to sell their premium service since it narrows the gap above what you can get for free.
- Stop worrying and leave it to us: Well, no. There are several balls you still have to keep your eye on. You wouldn’t think a tiny operation like my blog mailing list would be very concerned with integrations and so on. Actually, I have a website, a blog (which has now been integrated with my website thanks to the magic of WordPress), an email list, and an auto-responder that can send out multiple emails at once. Not to mention video and other content-generation capabilities, and repositories for all those files and backups for those too. Not all of them work smoothly with the others. I could spend $300+ per month putting everything on one platform, but it’s a lot of money to support what I do, and it makes it really questionable as to who actually owns all the content if I am totally dependent on the provider to host it all. That’s just me. Now consider the orders of magnitude of complexity of a proper organization’s IT operations! Your cloud objects will also need to interface and interact with each other, and you don’t have the luxury of doing it in your on-premises center where you can win the day (eventually) through some combination of brute force, trial and error, and duct tape.
- Your data is still your data: Most cloud providers cheerfully proclaim that any time you want to switch services they will give you your data back. Set aside the matter of whether you have the ability to host the volumes of data on your own assets (and with videos and images, even my little operation is ballooning towards a terabyte). You still have to consider the form in which the provider will deliver that data, and whether it really is complete. Linked-In, for instance, actually provides only a limited subset of your contact data, which is fair enough when they are the ones who provided the contacts to you. Google Contacts, by contrast, if they are able to identify your contact by the email address as being already in their database, considers all the related information to be theirs, even if they got it by scraping it from what you entered in about that contact. For enterprise-scale SaaS systems, you’ll get your data broken up in flat files representing whatever table format they use for their system design; it’ll be up to you to figure out how to put those dozens or hundreds of tables back into a usable form.
- You don’t need the skills, leave that to us. Not exactly. True, most organizations can’t afford to cover all the technical knowledge required, even when they could spread the knowledge across a whole IT staff, so it does make sense to take advantage of a cloud provider’s even-larger dedicated operation and them maintain currency. That doesn’t mean you are out of the woods. You may not need the detailed skills of being a server administrator (but see the next bullet) but more than ever you will need IT management skills. If you thought that moving to the cloud would somehow alleviate the chaos in your environment due to a lack of governance, the cloud will simply allow you to get to a state of even more confusion faster. You’ll need a strong architecture group to figure out how your layers are dis-aggregated, where all the parts are, and how they need to be integrated. Your engineers will keep an eye on the market to make sure that you keep pointed in the right direction. Your configuration management team will still need to know how many licenses you have for what applications. Your contracts team will be busy managing new kinds of arrangements, your finance people will still be trying to make sure they have the money they need when it is needed, and your security team will have to figure out how to manage all this across environments that you cannot even see, much less control.
- Caveat to all that: you may still need the technicians too, depending on what exactly you bought. If you decide to “go to the cloud” bu enrolling in Amazon Web Services, when you open up your web link you’ll be presented with a list of hardware and software that you must choose from and configure. True, you wont need to “administer” the server, meaning putting on patches and so forth, but you will have to know what to do when the operating system patch that is applied by the hosting service in the background makes your application stop working.
OK, so maybe I am just a techno-klutz. The truth is, even if you work in the IT business, you probably are too. Nobody can be expert on everything at once. Didn’t we just say that is one of the points of moving to the cloud? But in the cloud, you’re still going to need most of the experts you already have. The difference is that you won’t be able to kluge your way through problems or negotiate to get your peers to look the other way.
Having said all that, you can’t avoid dealing with cloud operations, nor should you. You can indeed save money, time, payroll, and many other distractions and expenses. In many cases you simply won’t be able to deliver a solution any other way (unless you custom code it yourself). Just don’t think that it’s all smooth sailing.
Have you got some happy stories to share? Or some hard-won lessons learned? Or maybe you disagree with my thinking on this. Feel free to leave a reply below, or share this out on social media to others with this interest or expertise.
Ph.D, PgMP, LSS Master Black Belt, RTBMA
Decision Integration LLC